Menu
Implement a plan from a previous session in an isolated worktree. Only use when the user explicitly invokes /implement. Do NOT auto-trigger.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
| name | implement |
| description | Implement a plan from a previous session in an isolated worktree. Only use when the user explicitly invokes /implement. Do NOT auto-trigger. |
Implement a plan (typically drafted in a prior session) in a git worktree, then open a PR.
Dirty branch check. Run git status --porcelain. If there are uncommitted changes, show them and ask the user whether to stash, commit, or abort before proceeding.
Locate the plan. Ask the user to paste it or point to a file/URL.
Notebook entry. Spawn a memory sub-agent (background, Sonnet) to create a notebook entry before starting work. To get the prompt template:
CONFIG_REPO=$(readlink ~/.claude/CLAUDE.md | xargs dirname | xargs dirname)
Read $CONFIG_REPO/templates/memory-agent-prompt.md and use it as the agent prompt. The entry should record the plan being implemented and will be updated at completion.
Enter worktree. Use the EnterWorktree tool.
Use Explore subagents (in parallel where independent) to understand the files and patterns the plan touches. In addition to plan-tagged files, have explorers search for other code paths that could plausibly be affected and report unexpected interactions or hidden dependencies back to the main agent. Read key files yourself for anything central to the change.
Review the plan against actual code. Identify:
Discuss major deviations with the user. If you identify a simpler high-level path to the same goal, discuss that alternative with the user before implementation. Minor improvements (better naming, simpler logic, obvious fixes) can proceed without asking.
Even if the plan didn't mention it, check whether these need updating:
CLAUDE.md (project instructions)config/prompts/.edn configs:docs / :short-docs metadatanotebook/writeup/ documentsCreate a PR using gh pr create. Include:
After the PR is created:
notebook/TODO.md and check whether the completed work corresponds to any to-do item. If so, move that item to notebook/DONE.md (with a completion date and reference to the PR).Navigate and develop the Spell source code. Use when modifying Spell internals, investigating evaluator/runtime/provider behavior, finding the right source files, updating tests, or explaining how the implementation fits together.
Set up Spell from a fresh checkout and run the first smoke tests or examples. Use when a user asks to install Spell, get started, verify the CLI, configure provider credentials, choose a first example, or try Spell for the first time.
Use Spell from the CLI, the Clojure API, and the Python benchmark adapter. Use when writing commands, embedding Spell in Clojure, calling Spell from benchmark Python code, or checking current API vocabulary.
Configure custom Spell agent and model profiles. Use when creating or modifying .agent.edn files, choosing namespaces, changing system prompts, wiring sub-agents, selecting provider-prefixed models, or doing agent smoke tests.
Navigate and develop the Spell source code. Use when modifying Spell internals, investigating evaluator/runtime/provider behavior, finding the right source files, updating tests, or explaining how the implementation fits together.
Set up Spell from a fresh checkout and verify prerequisites, CLI smoke tests, and available model provider authorization. Use when helping a user get to a runnable checkout, configure API provider access, or try the first Spell example.